Metabolomic analysis reveals trimethylamine N-oxide as a biomarker for poor outcome of severe spontaneous intracerebral hemorrhage patients receiving surgical treatment

Background and objectives Patients suffering from severe spontaneous intracerebral hemorrhage (SSICH) are at high risk of cardiocerebrovascular diseases postoperatively, which hugely affect patients’ long-term outcomes. Metabolic features could reflect the pathological change of the cardiocerebrovascular system and might serve as biomarkers for evaluating the risk of poor outcomes in SSICH patients. The current study aimed to find the early-warning biomarkers for poor outcomes in SSICH patients after surgery. Methods Severe spontaneous intracerebral hemorrhage patients receiving surgical treatment from a referring hospital were prospectively included and formed the primary cohort after propensity score matching. The primary outcome is poor 180 days after hemorrhage (modified Rankin scale ≥4). Metabolomics analysis on 3-, 7-, and 30-day serum and cerebrospinal fluid samples after surgery revealed the dysregulated metabolites of SSICH patients within the primary cohort. Within the validation cohort of SSICH patients receiving surgical treatment from a multicenter, prospective cohort, dysregulated metabolites were validated and evaluated to see whether they could serve as biomarkers for 180-day poor outcomes by area under the curve (AUC). Results The primary cohort included 20 SSICH patients with good 180-day outcome and 20 with poor outcome. Untargeted metabolomics analysis found 25 co-dysregulated metabolites, including trimethylamine N-oxide (TMAO), among 3-day, 7-day, and 30-day metabolism features between SSICH patients with poor outcome and good outcome after surgery. A good correlation was found in TMAO between serum and cerebrospinal fluid on 3rd day after surgery. Based on the validation cohort of 794 SSICH patients (147 patients had 180-day poor outcome), the targeted metabolomics analysis revealed increasing TMAO on 3rd day after surgery as a risk factor of poor outcomes (odds ratio, 4.7; 95%CI, 3.6–6.2; p < 0.001), with a good predictive value (AUC, 0.81). Conclusion This study demonstrated increasing serum TMAO level as an early-warning biomarker for 180-day poor outcomes of SSICH patients receiving surgical treatment. Clinical trial registration Chinese Clinical Trial Registry, ChiCTR1900024406, https://www.chictr.org.cn/showproj.html?proj=40640 .